Methodologic Considerations in assessing Effects of Air Pollution on Human Health

Air pollution, especially particulate matter (PM), has been shown to be related to health of populations, specifically as measured by increased incidence of respiratory and cardiovascular conditions and premature death. However, the relationship among science, politics, and policy may be no clearer than in assessing the effect of the effect of environment on health. Understanding the extent to which changes in measures of air pollution can be associated with those of human health is dependent not only on measurement of air pollution, but also on important statistical issues inherent in analytic modeling and analysis. In this presentation, we discuss three important concerns. First, we examine the additional information provided by multiple cause of death coding (in addition to underlying cause of death) for respiratory and cardiovascular disease. Second, we present an innovative method for assigning distance between location of death and air quality monitoring source, and we show the sensitivity of mortality estimates to this distance. Finally, we examine parameters associated with smoothing of weather data included in models of the relationship between PM and mortality for selected geographical regions in the United States. Results should enable analysts in the field of health and the environment to more completely understand the effect of air pollution and inform policy in this arena.